4.1. Survey of genetic elements introduced into approved transgenic
The structural genes and the respective regulatory sequences (promoters
and terminators) which have been used as transgenes are summarised
in the following sections. Depending on their frequency of use
in approved genetically modified crops and the probability of
a 'natural' occurrence of these sequences in food, sequences within
single genetic elements are applicable for the screening for GMO-food.
Genetic elements that have been used in isolated cases may allow
specific detection of the given product for as long as the respective
element is not employed in further approved products. Thus, methods
based on the detection of sequences within a single genetic element
will in the long run be better suited for screening purposes.
4.1.1 Survey of the structural genes used
Almost 35 distinct structural genes (including variants) have
been used for the generation of the currently approved transgenic
crops (Structural genes introduced into approved transgenic crops).
Some of these genes such as accd, accS, sam-k and some genes coding viral
coat proteins occur only in a single genetically engineered product.
Therefore, the identification of sequences of one of these genes in food
would represent a product-specific detection method provided the
actual sample did not contain the natural sources of these sequences
(e.g. from bacteriophages or plant viruses).
The most frequently used transgene is the nptII gene, originally
isolated from the bacterial transposon 5. The nptII gene has been
introduced into 17 out of the 28 approved agricultural crops.
In 16 products it functioned as a marker gene under the control
of a eucaryotic promoter; thus, nptII sequences seem to be well
suited for screening purposes. It should be noted, however, that
nptII occurs frequently in bacteria found in the environment (Smalla
et al., 1993; Redenbaugh et al., 1994). The presence of these
naturally occurring bacteria in a sample may, therefore, lead
to a false-positive result. Future transgenic crops are expected
to contain fewer or no marker genes in the final products since
marker-free insertion techniques or methods to eliminate marker
genes from transgenic plants (for review see Yolder and Goldbrough,
1994 and references in Niederhauser et al., 1996) and microorganisms
(Sanchis et al., 1997) are already available.
Other structural genes have been employed less frequently. From
the 28 approved crops, variants of the -endotoxin gene from Bacillus
thuringiensis or of the bar gene originally isolated from
Streptomyces hygroscopicus are found in 6 products each.
Variants of the CP4 epsps gene from Agrobacterium, the
-lactamase gene and of the polygalacturonase gene have been introduced
in 3 to 5 products each (Figure 4). In these cases various factors
may have to be assessed to judge the applicability of DNA-based
detection methods: (i) the presence of the transgene in the respective
transformation event (line); (ii) the 'completeness' of the respective
sequence (incompletely transferred, or 'truncated' or 'altered'
versions of genes may be present); (iii) the use of 'synthetic'
versions of genes that have an altered codon usage in order to
optimise gene expression in the host organism.
Figure 4: Number of occurrences of the most frequently
used transgenes introduced into the currently approved genetically
engineered agricultural crops (in total 28 distinct products were
approved; see text). In some cases (indicated by asterisks) distinct
variants of genes or 'synthetic' versions were used. (See also
Structural genes introduced into approved transgenic crops.)
Both the gene encoding for barnase and the aad gene are present
in three products each, whereas a synthetic version of the pat
gene can be found in four products. Seventeen other genes are
present in one or two of 17 different products.
Sequences from structural genes originating from homologous sources
(up to now mostly antisense constructs in tomatoes) are suitable
for the detection of GMOs only if certain prerequisites are fulfilled.
When the coding sequences occur in the (copies of the) transgene
and also in the naturally occurring copies of the gene, a clever
choice of primers for a PCR assay may allow discrimination of
the amplification products of the native gene and the transgene
by the length of the amplified fragment. This can be achieved
if the two primers bind to sequences on the chromosomal gene that
are situated on different (normally adjacent) exons. Whereas analysis
of conventional products would result in the amplification of
a single long fragment that includes the sequence of the intron
between the respective primer binding sites, analysis of the corresponding
genetically modified product would result in the appearance of
an additional, shorter amplification product lacking the intron
sequence (since transgenes originate from c-DNA sequences). A
description of this methodology was contained in the petition
for the genetically engineered tomato from Zeneca (Petition from
Zeneca for genetically modified tomatoes, 1995). However, such
a strategy requires not only the knowledge of the c-DNA sequence
(of the transgene) but also precise information about the intron-exon
boundaries of the chromosomal gene, which is not always available.